Speaker-independent style conversion for HMM-based expressive speech synthesis | IEEE Conference Publication | IEEE Xplore